VIA Interaction

Monday, November 17, 2008

Making sense of interaction

A couple of things caught my eye today, not unrelated.

In "What the Google phone stole from the iPhone", Slate's Farhad Manjoo makes the point that is is not what the iPhone and T-Mobile's new G1, which is based on Google's Android OS, do, it's how they do it:

"This sounds obvious: Doesn't every mobile phone company set out to create a usable interface? Spend a minute trying to navigate deep lists of drop-down menus on a Windows Mobile or BlackBerry device and you'll have your answer. Before the iPhone, phones were pretty to look at but a pain to use; the last blockbuster mobile phone, Motorola's RAZR, induced aneurysms when you tried to do anything but make a phone call. The iPhone changed all that: In the same way that the Mac proved that people want computers that can display calligraphy, the iPhone proved that people want phones that don't require a manual."

Steve Jobs is famous for trusting his own instincts or the instincts of his genius designers:

"Steve Jobs eschews focus groups. He likes to say that he doesn't believe in asking customers what they want; he prefers to build stuff in order to show customers what they want."

Which brings us to the second point. Lest it seem that Jobs's attitude stems only from arrogance, here's what usability expert Jakob Nielsen wrote today:

"In any case, what users want and what users need are two different things, which is why it's long been a primary usability guideline to watch what users do, rather than listen to what they say."

We do a lot of work with large, complex Web sites. We've done enough usability testing (formally and informally) to know that even the most successful designs need adjusting.

Unfortunately, working with large teams often forces us into "waterfall" projects, sequential development in which requirements are gathered, design follows, implementation comes next, and then the product is verified against the design and requirements. But not against whether it actually works for the target audience.

There's a big problem with that. As Nielsen put it, "requirement specifications are always wrong."

Always.

The answer is easy enough: Interface design requires iteration. Somewhere deep in the hideaways at Apple I am sure even Steve Jobs and Jonathan Ives have discarded some pretty good ideas when they turned out not to work at all in practice.

Selling this notion is difficult in a world where CMOs feel they must sign off on every design decision prior to implementation. That is almost certainly a recipe for failure. The real world collides with design theory every time. And it is the customers who pay.

posted by Tim Beidel at 11/17/2008 09:16:00 PM

---